STATISTICAL ALTERNATIVES FOR STUDYING COLLEGE-STUDENT RETENTION - A COMPARATIVE-ANALYSIS OF LOGIT, PROBIT, AND LINEAR-REGRESSION

While higher education researchers have long been concerned with the d evelopment and application of methods to adequately assess the impact of college on students, strong advances in statistical theory and comp utational practice have shifted this focus from the fundamental issues of research design to the application of appropriate statistics. This study focuses on the practical implications of applying logistic regr ession, probit analysis, and linear regression to the problem of predi cting college student retention. Rather than simply assuming that one technique is analytically superior to others based on theoretical grou nds, this study explores how these techniques compare in predicting st udent retention using data provided by registrars from a national samp le of colleges and universities. Results indicate that despite the the oretical advantages offered by logistic regression and probit analysis , there is little practical difference between either of these two tec hniques and more traditional linear regression.